We describe a statistical approach for modeling dialogue acts in conversational speech, i.e., speec-act-like

Identifying whether an utterance is a statement, question, greeting, and so forth is integral to effective automatic understanding of natural dialog. Little is known, however, about how such dialog acts (DAs) can be automatically classified in truly natural conversation. This study asks whether current approaches, which use mainly word information, could be improved by adding prosodic information. The study is based on more than 1000 conversations from the Switchboard corpus. DAs were handannotated, and prosodic features (duration, pause, F0, energy, and speaking rate) were automatically extracted for each DA. In training, decision trees based on these features were inferred

We describe a new corpus of over 180,000 handannotated dialog act tags and accompanying adjacency pair annotations for roughly 72 hours of speech from 75 naturally-occurring meetings. We provide a brief summary of the annotation system and labeling procedure, inter-annotator reliability statistics, overall distributional statistics, a description of auxiliary files distributed with the corpus, and information on how to obtain the data.

We explore the two related tasks of dialog act (DA) segmentation and DA classification for speech from the ICSI Meeting Corpus. We employ simple lexical and prosodic knowledge sources, and compare results for human-transcribed versus automatically recognized words. Since there is little previous work on DA segmentation and classification in the meeting domain, our study provides baseline performance rates for both tasks. We introduce a range of metrics for use in evaluation, each of which measures different aspects of interest. Results show that both tasks are difficult, particularly for a fully automatic system. We find that a very simple prosodic model aids performance over lexical information alone, especially for segmentation. Both tasks, but particularly word-based segmentation, are degraded by word recognition errors. Finally, while classification results for meeting data show some similarities to previous results for telephone conversations, findings also suggest a potential difference with respect to the effect of modeling DA context.

This paper describes the application of the PARADISE evaluation framework to the corpus of 662 human-computer dialogues collected in the June 2000 Darpa Communicator data collection. We describe results based on the standard logfile metrics as well as results based on additional qualitative metrics derived using the DATE dialogue act tagging scheme. We show that performance models derived via using the standard metrics can account for 37% of the variance in user satisfaction, and that the addition of DATE metrics improved the models by an absolute 5%.

This paper describes a dialogue act tagging scheme developed for the purpose of providing finer-grained quantitative dialogue metrics for comparing and evaluating DARPA COMMUNICATOR spoken dialogue systems. We show that these dialogue act metrics can be used to quantify the amount of effort spent in a dialogue maintaining the channel of communication or, establishing the frame for communication, as opposed to actually carrying out the travel planning task that the system is designed to support. We show that the use of these metrics results in a 7% improvement in the fit in models of user satisfaction. We suggest that dialogue act metrics can ultimately support more focused qualitative analysis of the role of various dialogue strategy parameters, e.g. initiative, across dialogue systems, thus clarifying what development paths might be feasible for enhancing user satisfaction in future versions of these systems. 1.

We investigate whether automatically extracted prosodic features can serve as cues to dialog acts (DAs) in naturallyoccurring meetings. We focus on the classification of four short DAs, all of which can be conveyed by the same words. DAs were hand-labeled based on the discourse context. Results for classifiers trained on automatically extracted prosodic features show significant associations with DAs in unseen test data. Furthermore, the specific features used depend on the classification task at hand. Results shed light on the relationship between discourse function and prosody, and could be used to aid automatic processing for natural dialog understanding.

This paper presents a system which automatically generates shallow semantic frame structures for conversational speech in unrestricted domains. We argue that such shallow semantic representations can indeed be generated with a minimum amount of linguistic knowledge engineering and without having to explicitly construct a semantic knowledge base. The system is designed to be robust to deal with the problems of speech dysfluencies, ungrammaticalities, and imperfect speech recognition. Initial results on speech transcripts are promising in that correct mappings could be identified in 21% of the clauses of a test set (resp. 44% of this test set where ungrammatical or verb-less clauses were removed). 1 Introduction In syntactic and semantic analysis of spontaneous speech, little research has been done with regard to dealing with language in unrestricted domains. There are several reasons why so far an in-depth analysis of this type of language data has been considered prohibitively hard: ...

In this work we investigate the use of support vector machines (SVMs) and discriminative learning techniques on the task of automatic classification of dialogue acts (DAs) from prosodic cues. We implement and test these classifiers on solving an 8DA classification task on the Spanish CallHome database and report preliminary recognition rates of 47.3% with respect to a 20.4% chance-level rate, which represents an improvement over previously reported work using decision trees and neural network classifiers. Although prosodic cues alone may not suffice for robust classification of DAs, we report results that suggest that SVMs offer an interesting alternative to previously explored models, and should be further explored to improve the contribution of prosodic models to the classification task.

This paper describes a 3-level manual discourse coding scheme that we have devised for manual tagging of the CallHome Spanish (CHS) and CallFriend Spanish (CFS) databases used in the CLARITY project. The goal of CLARITY is to explore the use of discourse structure in understanding conversational speech. The project combines empirical methods for dialogue processing with state-of-the art LVCSR (using the JANUS recognizer). The three levels of the coding scheme are (1) a speech act level consisting of a tag set extended from DAMSL and Switchboard; (2) dialogue game level defined by initiative and intention; and (3) an activity level defined within topic units. The manually tagged dialogues are used to train automatic classifiers. We present preliminary results for statement categorization, and give an in-progress report of automatic speech act classification and topic boundary identification. 1. INTRODUCTION To appear in: International Conference on Spoken Language Processing (ICSLP'98...